Event, descriptive, and evaluative components of the narrative in professional and AI-generated sports commentary: a comparative linguonarratological analysis

The aim of the study is to identify and describe systemic differences in the narrative organization of texts produced by professional sports commentators and generative artificial intelligence through comparative linguonarratological analysis. The subject of the research is the narrative strategies reflected in the texts produced by sports commentators and generative AI systems. The relevance of this study is determined by the rapid implementation of generative AI systems in sports journalism and media commentary, which presents fundamentally new problems for contemporary language science related to the nature of automatically generated text, its narrative organization, and evaluative potential. Furthermore, sports commentary, as one of the genres of media discourse, represents a unique communicative environment in which the narrative subjectivity of the speaker acts as a systemic feature of the text, forming its pragmatic potential, identificational function, and emotional-persuasive impact on the audience. Identifying the specifics of the aforementioned characteristics is of interest to both language theory and media communication practice. The main methods of this work are general scientific methods (description, comparison, and systematization of the analyzed material) and linguistic methods (linguonarratological method, discourse analysis, contextual analysis). Additionally, elements of quantitative analysis were used to verify the identified patterns at a statistical level. The scientific novelty of the research lies in the fact that, for the first time in domestic linguistics, a comparison is made between professional sports commentary and AI-generated text using an integrative methodology that combines the tools of comparative linguonarratology, linguoaxiology, and discursive linguistics. Conclusions: texts generated by the ChatGPT language model (Open AI) are characterized by structural hypercorrection. At the same time, texts of professional commentators exhibit a reduced scheme: the abstract and code are minimal, orientation is realized fractionally and embedded in complication, and the resolution is merged with evaluation. In the sports commentary of professional journalists, evaluation is evenly distributed throughout the narrative, activated with a rich set of means. In AI-generated text, evaluation mainly concentrates in the resolution and code. Professional texts demonstrate an event-evaluative and dialogically event-evaluative configuration. AI texts are characterized by an event-conclusive and event-protocol configuration.